Rate limits, caps, and cost control

Conversation caps, origin checks, Shopify timeouts, and how to keep AI spend predictable.

Formo enforces origin checks, rate limits, and monthly conversation caps. You still set a spend cap at your AI provider.

Conversation caps#

Standard: 2,000 conversations/month
Plus: 10,000 conversations/month

Exceeding the cap switches the widget to contact capture. There is no overage bill from Formo. Counters reset on the 1st of the month (UTC). You get an email at 80% and at 100%.

Origin#

site_… in the script tag is public. Only hosts on the site’s allowed-domains list can load config or chat. A copied tag on another domain receives 403.

Rate limits#

20 messages/minute per visitor hash
300 messages/minute per site
40 new conversations/hour per visitor hash

Shopify#

Each Admin API call uses a 10 second timeout and two retries, and returns explicit not-found rather than an empty success. Prefer asking for one order number over scanning the whole shop.

AI spend#

Pin a model and a fast model on AI Provider.
Set a hard limit in the provider console.
Watch estimated monthly cost on that page (from last 30 days of tokens).
Stay in Dry Run until Activity is clean.

Formo never bills per message. If you need more sites or a higher cap, Plus is the one-time upgrade (£299, 3 sites, 10,000 conversations/month).